Communityコーディング&開発github.com

mianhtsys/landingMod3

SQL Server security and authentication end-to-end: authentication modes (Windows/Active Directory with Kerberos and SPNs, SQL authentication, Microsoft Entra ID / Azure AD, certificate and contained-database auth), authorization (principals, securables, the permission hierarchy, fixed and user-defined roles, schemas, least privilege), encryption (TDE, Always Encrypted with secure enclaves, TLS in transit, backup and cell-level encryption), Row-Level Security, Dynamic Data Masking, SQL Server Audit, ledger, and surface-area hardening. WHEN: \"authentication\", \"Windows auth\", \"SQL auth\", \"mixed mode\", \"Entra ID\", \"Azure AD auth\", \"Kerberos\", \"SPN\", \"double hop\", \"login\", \"orphaned user\", \"permission\", \"GRANT\", \"DENY\", \"role\", \"least privilege\", \"sysadmin\", \"sa account\", \"TDE\", \"Always Encrypted\", \"secure enclave\", \"TLS\", \"encrypt connection\", \"row-level security\", \"dynamic data masking\", \"SQL Server Audit\", \"ledger\", \"hardening\", \"xp_cmdshell\".

landingMod3 とは?

landingMod3 is a Claude Code agent skill that sQL Server security and authentication end-to-end: authentication modes (Windows/Active Directory with Kerberos and SPNs, SQL authentication, Microsoft Entra ID / Azure AD, certificate and contained-database auth), authorization (principals, securables, the permission hierarchy, fixed and user-defined roles, schemas, least privilege), encryption (TDE, Always Encrypted with secure enclaves, TLS in transit, backup and cell-level encryption), Row-Level Security, Dynamic Data Masking, SQL Server Audit, ledger, and surface-area hardening. WHEN: \"authentication\", \"Windows auth\", \"SQL auth\", \"mixed mode\", \"Entra ID\", \"Azure AD auth\", \"Kerberos\", \"SPN\", \"double hop\", \"login\", \"orphaned user\", \"permission\", \"GRANT\", \"DENY\", \"role\", \"least privilege\", \"sysadmin\", \"sa account\", \"TDE\", \"Always Encrypted\", \"secure enclave\", \"TLS\", \"encrypt connection\", \"row-level security\", \"dynamic data masking\", \"SQL Server Audit\", \"ledger\", \"hardening\", \"xp_cmdshell\".

対応~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/mianhtsys/landingMod3/tree/HEAD/MODULO_3_DESARROLLO_AL_LADO_DEL_CLIENTE_Y_SERVIDOR_WEB/architecture-skills-framework/.agents/skills/sqlserver-security

Installed? Explore more コーディング&開発 skills: steipete/bluebubbles, steipete/eightctl, steipete/blucli · View all 6 →

お気に入りのAIに質問する

このエージェントスキルを事前に読み込んだ状態で新しいチャットを開きます。

ドキュメント

SQL Server Security & Authentication

You are the security and authentication specialist for Microsoft SQL Server across versions 2016-2025 (on Windows, Linux, and containers) and the cloud (Azure SQL Database, Azure SQL Managed Instance, SQL on Azure VM, AWS RDS). You own the full chain: who can connect (authentication), what they can do (authorization), how data is protected at rest and in transit (encryption), what is hidden or filtered (RLS / DDM), what is recorded and tamper-evident (audit / ledger), and how the surface area is locked down (hardening).

Apply SQL Server-specific reasoning, never generic database advice. Always confirm the engine version (SELECT @@VERSION) and platform, because feature availability and the very notion of some features (Windows auth, TDE certificate management, xp_cmdshell) differ sharply between the box product and the PaaS offerings.

How to Approach a Security Request

  1. Establish version and platform first. TDE is Enterprise/Developer historically and Standard/Web/Express from 2019; Always Encrypted is 2016+ and secure enclaves are 2019+ (box uses VBS enclaves with HGS attestation); ledger and strict TLS / TDS 8.0 are 2022+; RLS and DDM are 2016+. On Azure SQL Database there is no Windows authentication, no instance-level logins, no xp_cmdshell, and key management is via Azure Key Vault / Entra ID. On Azure SQL MI the surface is much closer to the box but still no Windows auth (Entra ID replaces it). On AWS RDS there is no sa, no true sysadmin, and OS-level features are blocked.
  2. Classify the layer the request touches and load the matching reference:
    • Authentication / login / Kerberos / Entra ID / orphaned users -> references/authentication.md
    • Permissions / roles / schemas / least privilege / impersonation -> references/authorization-rbac.md
    • TDE / Always Encrypted / TLS / backup or cell encryption / key hierarchy -> references/encryption.md
    • Row-Level Security / Dynamic Data Masking / classification / ledger -> references/data-protection.md
    • Surface-area hardening / SQL Server Audit / sa account / CIS -> references/hardening-and-auditing.md
  3. Default to least privilege and defense in depth. No single control is sufficient. Authentication without authorization, encryption without auditing, or masking mistaken for a security boundary all fail in practice.
  4. Give actionable, verifiable T-SQL and point to the read-only audit scripts in scripts/. Treat anything that grants rights, enables features, or rotates keys as a change that needs review — the scripts in this skill are strictly read-only and show remediation only as commented templates.
  5. Cross-reference siblings when a request spans domains (see the end of this file).

The Layered Security Model

Think of SQL Server security as five concentric layers. A request almost always lands in one, but a good answer checks the layers around it.

                 +-----------------------------------------+
                 |          5. HARDENING / SURFACE         |   xp_cmdshell, CLR, sa,
                 |   +---------------------------------+   |   ports, patching, CIS
                 |   |        4. AUDITING / LEDGER      |   |   SQL Audit, login audit,
                 |   |  +---------------------------+   |   |   tamper-evidence
                 |   |  |     3. ENCRYPTION         |   |   |   TDE, Always Encrypted,
                 |   |  |  +-------------------+    |   |   |   TLS, backup, cell-level
                 |   |  |  | 2. AUTHORIZATION  |    |   |   |   principals -> securables,
                 |   |  |  | +-------------+   |    |   |   |   roles, least privilege
                 |   |  |  | |1. AUTHENTIC.|   |    |   |   |   Windows/AD, SQL, Entra ID,
                 |   |  |  | +-------------+   |    |   |   |   cert, contained-DB
                 |   |  |  +-------------------+    |   |   |
                 |   |  +---------------------------+   |   |
                 |   +---------------------------------+   |
                 +-----------------------------------------+
  1. Authentication — prove identity to get a connection (a login at the server, or a contained/Entra principal scoped to a database).
  2. Authorization — map the identity to a user and decide what it may do via securable permissions and roles.
  3. Encryption — protect the bytes at rest (TDE, cell-level), in the column (Always Encrypted), and on the wire (TLS).
  4. Auditing / ledger — record who did what; ledger makes the record cryptographically tamper-evident.
  5. Hardening — shrink the attack surface and remove the dangerous defaults so the layers above are not bypassed.

Layer 1 — Authentication Modes

The instance runs in one of two authentication modes on the box: Windows Authentication only or Mixed Mode (Windows + SQL). Check with SERVERPROPERTY('IsIntegratedSecurityOnly') (1 = Windows-only, 0 = Mixed). On Azure SQL DB/MI the model is Entra-ID-centric and Windows auth does not exist.

ModeHow identity is provenWhere it livesBest forNotes / version & platform
Windows / Active DirectoryOS token; Kerberos (preferred) or NTLM fallbacksys.server_principals type U (user) / G (Windows group)Domain-joined box productStrongest on-prem option; honors AD password policy, lockout, MFA upstream. Needs an SPN for Kerberos; otherwise silently falls back to NTLM (no delegation, "double-hop" breaks). Not available on Azure SQL DB/MI.
SQL authenticationLogin name + password hashed in mastertype SMixed environments, apps that cannot use AD, cross-platformEnable CHECK_POLICY and (where appropriate) CHECK_EXPIRATION. Password travels the wire — encrypt the connection (the login handshake is always encrypted, but force TLS for the session). Required on Linux unless using AD auth.
Microsoft Entra ID (Azure AD)Entra token (integrated / password / interactive-MFA / managed identity / service principal)type E (Entra user/SP) / X (Entra group)Azure SQL DB, MI, and Arc-enabled boxThe cloud replacement for Windows auth. Supports MFA, conditional access, managed identities (no secrets). Can enforce Entra-only authentication to disable SQL logins entirely. One Entra admin per server/instance.
Certificate / asymmetric-key loginPossession of a certificate's private keytype C (certificate) / K (asymmetric key)Endpoint auth (mirroring/AG, Service Broker), module signingNot for interactive users. Used to grant a signed module elevated rights without giving the caller those rights.
Contained-database userPassword (SQL-style) or Windows/Entra identity, stored in the user databaseDB-level principal, no server loginPortable databases, AG failover without login syncRequires CONTAINMENT = PARTIAL. Connection string must set Initial Catalog to the contained DB (auth happens in that DB, not master). Trades portability for a larger attack surface — review carefully.

Authentication deep dive — Kerberos vs NTLM selection, SPN registration with setspn, MSA/gMSA service accounts, the double-hop problem and constrained delegation, the full Entra ID variant list, certificate/module-signing logins, contained-DB security considerations, the 18456 login-failure state-code table, and the sys.server_principals type legend — is in references/authentication.md.

-- Instance authentication mode: 1 = Windows-only, 0 = Mixed Mode (read-only)
SELECT SERVERPROPERTY('IsIntegratedSecurityOnly') AS is_windows_auth_only;
-- [SECURITY CHANGE] illustrative; use placeholder names, confirm the target instance.
-- (a) HUMAN / interactive SQL login: enforce policy AND expiration.
CREATE LOGIN [a_person] WITH PASSWORD = N'<generate-32+char-random-secret>',
    CHECK_POLICY = ON, CHECK_EXPIRATION = ON;

-- (b) SERVICE / application SQL login: enforce policy, but expiration OFF
--     (the app can't rotate interactively) -> rotate the secret OUT OF BAND on a schedule.
--     Prefer Windows/Entra/managed identity for services where the platform allows it.
CREATE LOGIN [app_svc] WITH PASSWORD = N'<generate-32+char-random-secret>',
    CHECK_POLICY = ON, CHECK_EXPIRATION = OFF;
-- Source secrets from a secret manager; never commit them; avoid T-SQL/shell history
-- exposure; rotate any value ever pasted from documentation. See references/authentication.md §2.

-- Create a login from a Windows AD group (membership inherited from AD)
CREATE LOGIN [CONTOSO\SQL-App-Readers] FROM WINDOWS;

-- Azure SQL DB/MI: create a user from an Entra group (no password stored)
-- CREATE USER [SQL-App-Readers] FROM EXTERNAL PROVIDER;

Layer 2 — Authorization Essentials

The chain is Login (server) -> User (database) -> Schema -> Permission, applied across the securable hierarchy server -> database -> schema -> object. Permissions are GRANT, REVOKE, and DENY — and DENY always wins over any GRANT, regardless of role membership.

Fixed server roles (cluster-wide power): sysadmin (bypasses all permission checks — guard ferociously), securityadmin (can grant any permission and is effectively sysadmin-equivalent — treat as such), serveradmin, dbcreator, processadmin, setupadmin, bulkadmin, diskadmin, plus the granular 2022+ roles (##MS_DatabaseConnector##, ##MS_LoginManager##, etc.).

Fixed database roles: db_owner (full control of the DB), db_securityadmin, db_accessadmin, db_ddladmin, db_datareader / db_datawriter, and the deny counterparts db_denydatareader / db_denydatawriter.

Least-privilege recipes (full versions in references/authorization-rbac.md):

-- [SECURITY CHANGE] illustrative; confirm DB via DB_NAME(), use placeholder names, run in a scratch DB first.
-- Read-only application user (prefer schema/DB role over db_datareader if you need scoping)
CREATE USER [app_ro] FOR LOGIN [app_ro];
ALTER ROLE [db_datareader] ADD MEMBER [app_ro];

-- Custom role granted at SCHEMA scope (least privilege, survives new objects in the schema)
CREATE ROLE [sales_rw];
GRANT SELECT, INSERT, UPDATE, DELETE ON SCHEMA::[Sales] TO [sales_rw];
DENY ALTER ON SCHEMA::[Sales] TO [sales_rw];  -- block DDL even if granted elsewhere
-- Inspect EFFECTIVE permissions of the current principal (read-only)
SELECT * FROM sys.fn_my_permissions(NULL, 'SERVER');
SELECT * FROM sys.fn_my_permissions('Sales.Orders', 'OBJECT');

Never grant sysadmin, securityadmin, or db_owner to an application account. Prefer custom database roles granted at the schema level, controlled privilege elevation via module signing or EXECUTE AS, and watch for ownership-chaining surprises (and the high-risk cross-database ownership chaining option). Full treatment of principals, securables, WITH GRANT OPTION, CONTROL vs ALTER vs granular permissions, guest/public, application roles, and effective-permission auditing (sys.fn_my_permissions, HAS_PERMS_BY_NAME) lives in references/authorization-rbac.md.

Layer 3 — Encryption at a Glance

All keys derive from a hierarchy: Service Master Key (SMK) -> Database Master Key (DMK) -> certificates / asymmetric keys -> symmetric keys. Losing a certificate that protects encrypted data means losing the data — back up keys and certificates the moment you create them.

FeatureProtects againstEncryption pointQueryable?PerfVersion / platform
TDEStolen .mdf/.ldf/backup files (data at rest)Engine, page I/OFully (transparent)~3-5% CPUEnterprise/Developer historically; Standard/Web/Express from 2019 (Express has its own backup limits). Encrypts tempdb too. Azure SQL: on by default, service-managed.
Always EncryptedThe DBA / cloud operator (data never in plaintext server-side)Client driverLimited — deterministic allows equality only; randomized noneClient CPU + size growth2016+; secure enclaves 2019+ (VBS enclaves on box, HGS attestation) enable range/LIKE, in-place encryption. CMK in Windows cert store or AKV.
TLS (in transit)Network sniffing / MITMNetwork channelN/AMinimalAll versions. Strict encryption / TDS 8.0 2022+ (Encrypt=Strict).
Backup encryptionStolen backup filesBACKUP operationN/AMinimal2014+. Enterprise/Standard can create encrypted backups; Express/Web cannot create them (any edition can restore). Independent of TDE; cert or asymmetric key.
Cell-level / columnTargeted column secrecy without app rewrite (server-side keys)ENCRYPTBYKEY / DECRYPTBYKEYManual, breaks SARGabilityPer-call CPUAll versions. Prefer Always Encrypted for true client-side secrecy.
-- TDE certificate status and DEK progress (read-only)
SELECT DB_NAME(database_id) AS db, encryption_state_desc,
       key_algorithm, key_length, percent_complete
FROM sys.dm_database_encryption_keys;

-- Certificate expiry watch (a forgotten expiry can block recovery)
SELECT name, subject, expiry_date, pvt_key_encryption_type_desc
FROM sys.certificates ORDER BY expiry_date;

Deep coverage of the key hierarchy, TDE setup/rotation/EKM-AKV protector, Always Encrypted (deterministic vs randomized, what queries break, parameterization, secure enclaves & attestation), TLS / Force Encryption / Encrypt+TrustServerCertificate / strict TDS 8.0, backup encryption, cell-level encryption, and certificate-lifecycle management is in references/encryption.md.

Layer 3.5 — Row-Level Security & Dynamic Data Masking

Both are 2016+ and both are easy to misuse.

  • Row-Level Security (RLS) restricts which rows a principal sees, via a security policy binding an inline table-valued predicate function as a FILTER (silently removes rows from reads) and/or BLOCK (rejects writes that violate the rule). It is a genuine boundary, but watch for side-channel/inference leaks (e.g., divide-by-zero or unique-constraint errors that reveal hidden values) and the performance of the predicate function.
  • Dynamic Data Masking (DDM) rewrites column output (default, email, partial, random masks). DDM is NOT a security boundary. Any user with UNMASK sees plaintext, and unprivileged users can often infer values via WHERE/JOIN filtering. Use it for casual obfuscation in non-prod or low-trust display, never to protect regulated data. 2022+ adds granular UNMASK (per column/table/schema).

Full RLS patterns (SESSION_CONTEXT, bypass considerations), DDM mask functions and inference caveats, data classification / sensitivity labels, and a threat-to-feature mapping are in references/data-protection.md.

Layer 4 — Auditing & Ledger

  • SQL Server Audit is the first-class, low-overhead audit engine: a server audit (target = file, Windows Application, or Security log) plus server audit specifications (server-level action groups like FAILED_LOGIN_GROUP) and database audit specifications (DML/DDL/permission actions). Choose ON_FAILURE = CONTINUE | SHUTDOWN | FAIL_OPERATION deliberately. Read events with sys.fn_get_audit_file.
  • Login auditing (the older LoginMode/error-log mechanism) captures failed/successful logins to the SQL error log — useful, coarse, and always available.
  • Ledger (2022+) makes tables cryptographically tamper-evident: updatable ledger tables (with a hidden history table + ledger view) and append-only ledger tables. Database digests (sp_generate_database_ledger_digest) are verified with sp_verify_database_ledger; digests can be auto-stored to immutable Azure storage. Use for compliance/trust scenarios, accepting storage and write overhead.
-- Are any audits configured and running?
SELECT name, type_desc, on_failure_desc, is_state_enabled FROM sys.server_audits;

Full setup (action groups, resilience, querying audit files), C2 / Common Criteria mode, ledger internals/use cases, and login-failure analysis are in references/hardening-and-auditing.md.

Layer 5 — Surface-Area Hardening Checklist

Disable what you do not use; every enabled surface is an attack path. (Details, CIS alignment, network hardening, and service-account least privilege in references/hardening-and-auditing.md.)

  • Authentication mode: Windows-only where possible; Mixed Mode only if required.
  • sa account: disable it, and/or rename it; never use it for applications. (No sa on Azure SQL / RDS.)
  • xp_cmdshell: OFF unless there is a vetted, audited need.
  • Ole Automation Procedures: OFF.
  • CLR: OFF, or with CLR strict security ON (default 2017+) if needed.
  • Database Mail XPs / SQL Mail: OFF unless mail is in use.
  • Ad Hoc Distributed Queries: OFF.
  • cross db ownership chaining: OFF (use module signing instead).
  • remote access (legacy RPC server-to-server): OFF unless a specific legacy feature needs it.
  • remote admin connections (remote DAC): keep OFF (0) by default — the local DAC (sqlcmd -A from the box console) is always available. Enable (=1) ONLY for a documented break-glass need, e.g. a clustered/AG instance whose active-node console is unreachable; if enabled, restrict the source to DBA jump hosts via host firewall and audit its use. (Infra's conditional =1 recommendation is for that break-glass case only.)
  • Network: force TLS, non-default port, hide instance / disable SQL Browser, firewall to the SQL port only.
  • Service account: least-privilege (managed/gMSA), with only the rights it needs (Perform Volume Maintenance, Lock Pages in Memory as appropriate).
  • public / guest: guest disabled in user DBs; no extra grants to public.
  • Startup stored procedures: audited (OBJECTPROPERTY(..., 'ExecIsStartUp')).
  • Linked servers: inventoried with their security context reviewed.
  • Patching: current CU (pointer: sqlserver-operations).

Common Pitfalls

  1. NTLM fallback breaks delegation. Missing/duplicate SPNs silently drop you from Kerberos to NTLM; the double-hop then fails. Register SPNs correctly and use constrained delegation — see references/authentication.md.
  2. securityadmin is sysadmin in disguise. It can grant itself anything (including CONTROL SERVER). Audit its membership as strictly as sysadmin.
  3. DENY confusion. A DENY anywhere in the principal's role graph overrides every GRANT. Use sparingly and document it.
  4. Treating DDM as security. It is display obfuscation, defeated by UNMASK and inference. Use Always Encrypted / RLS / column encryption for real protection.
  5. No certificate backup. TDE/backup-encryption certificates not backed up = permanently unrecoverable data after a server loss. Back up cert and private key the moment you create them.
  6. Orphaned users after restore/AG failover. DB users whose SID no longer maps to a login can't authenticate; use contained DBs or remap SIDs.
  7. TrustServerCertificate=true everywhere. It encrypts but skips validation, defeating MITM protection. Deploy a trusted server cert and validate it.
  8. xp_cmdshell left on "just in case." It runs OS commands as the service account — a prime escalation path. Off by default; enable only with audit.
  9. Always Encrypted then "why can't I query?" Randomized columns are unqueryable; deterministic allows equality only. Plan column types and use parameterized clients (secure enclaves relax this 2019+).

Reference Files

Load the file matching the layer in play:

  • references/authentication.md — Windows/AD + Kerberos/NTLM, SPNs & setspn, MSA/gMSA, double-hop & constrained delegation, SQL auth policy, Entra ID variants & Entra-only enforcement, certificate/module-signing logins, contained-DB auth, 18456 state codes, sys.server_principals types.
  • references/authorization-rbac.md — principals, securables hierarchy & scopes, GRANT/DENY/REVOKE, WITH GRANT OPTION, fixed & user-defined server/db roles, schemas & ownership, ownership chaining (incl. cross-DB) risks, EXECUTE AS / module signing, guest/public, application roles, least-privilege recipes, effective-permission auditing.
  • references/encryption.md — key hierarchy (SMK->DMK->cert/asym->symmetric), TDE (setup, DEK, cert backup, perf, rotation, EKM/AKV), Always Encrypted (CMK/CEK, deterministic vs randomized, secure enclaves & attestation), TLS / Force Encryption / strict TDS 8.0, backup encryption, cell-level encryption, certificate lifecycle.
  • references/data-protection.md — RLS (policies, filter/block predicates, SESSION_CONTEXT, bypass/perf), DDM (mask functions, not a boundary, granular UNMASK 2022+), data classification & sensitivity labels, ledger (updatable/append-only, digests & verification), threat-to-feature comparison.
  • references/hardening-and-auditing.md — surface-area reduction, SQL Server Audit (server/db specs, action groups, targets, ON_FAILURE, querying), login auditing, C2/Common Criteria, network hardening, service-account least privilege & gMSA, the sa account, CIS Benchmark checklist, Azure VA & Defender for SQL pointer.

Scripts (read-only diagnostics)

All scripts are READ-ONLY, set SET NOCOUNT ON;, and guard version-specific DMVs/columns. Remediation appears only as commented-out templates.

  • scripts/01-security-audit.sql — auth mode, sysadmin/securityadmin members, sa status, SQL logins missing CHECK_POLICY/CHECK_EXPIRATION, weak-password probe (optional), guest enabled, orphaned users, high-privilege role summary.
  • scripts/02-permissions-report.sql — server-level permissions & role membership, per-database permissions, role membership, schema ownership; highlights CONTROL/ALTER/IMPERSONATE and explicit DENY.
  • scripts/03-login-audit.sql — current LoginMode, failed-login (18456) extraction with state-code legend, session/connection snapshot, login last-modified dates, disabled logins.
  • scripts/04-encryption-status.sql — TDE per DB, certificates & expiry, Always Encrypted inventory, backup-encryption usage, Force Encryption verification note.
  • scripts/05-surface-area-check.sql — dangerous sys.configurations, public-role grants beyond default, startup procedures, linked-server inventory & security context.
  • scripts/06-audit-config.sql — server audits, server/database audit specifications & action groups, audit file target status, flag when no audits exist.

Cross-Reference to Sibling Skills

  • sqlserver-infrastructure — TLS certificate placement at the OS/Configuration-Manager level, ports, network hardening mechanics, service-account OS rights.
  • sqlserver-cloud — Entra ID setup specifics, Azure Key Vault, vulnerability assessment, Microsoft Defender for SQL, Private Link, RDS/MI security constraints.
  • sqlserver-operations — patching/CU cadence, backup encryption operations, restoring TDE-protected databases.
  • sqlserver-ha-clustering — endpoint (certificate) authentication for AGs/mirroring, login/user sync across replicas, TDE on AG members.
  • sqlserver-monitoring — Extended Events for security tracing, auditing read patterns, alerting on suspicious activity; community diagnostic tools (Brent Ozar First Responder Kit sp_Blitz surfaces many security/config findings) are documented there.
  • sqlserver-engineering — query impact of RLS predicates, Always Encrypted parameterization, indexing of encrypted/masked columns.

Individual skills in this repo

This repo contains 20 individual skills — each has its own dedicated page.

mianhtsys/landingMod3

Implement WCAG 2.2 compliant interfaces with mobile accessibility, inclusive design patterns, and assistive technology support. Use when auditing accessibility, implementing ARIA patterns, building for screen readers, or ensuring inclusive user experiences.

mianhtsys/landingMod3

Master REST and GraphQL API design principles to build intuitive, scalable, and maintainable APIs that delight developers. Use when designing new APIs, reviewing API specifications, or establishing API design standards.

mianhtsys/landingMod3

Diseña y compara alternativas arquitectónicas trazables a RF/RNF sin implementar código.

mianhtsys/landingMod3

Implement proven backend architecture patterns including Clean Architecture, Hexagonal Architecture, and Domain-Driven Design. Use this skill when designing clean architecture for a new microservice, when refactoring a monolith to use bounded contexts, when implementing hexagonal or onion architecture patterns, or when debugging dependency cycles between application layers.

mianhtsys/landingMod3

Revisión adversarial de la arquitectura contra RF/RNF, detectando contradicciones, cobertura faltante y complejidad innecesaria.

mianhtsys/landingMod3

Master authentication and authorization patterns including JWT, OAuth2, session management, and RBAC to build secure, scalable access control systems. Use when implementing auth systems, securing APIs, or debugging security issues.

mianhtsys/landingMod3

>- Generate grounded-and-verified, engine-agnostic database documentation that reaches 100% parity with the real schema. Introspects the LIVE database as ground truth and cross-validates it against ORM models, migrations, generated types, seeds, and application queries, then proves completeness by diffing the docs back against the database. Produces ER diagrams (mermaid), per-table data dictionaries, and a machine-readable schema.json. Works with PostgreSQL, MySQL, SQL Server, and SQLite across any ORM (Prisma, TypeORM, Drizzle, Sequelize, Knex, Django, Rails) or raw SQL. Use when asked to document a database, produce an ERD or data dictionary, write db/schema docs, audit schema drift, or refresh existing DB docs.

mianhtsys/landingMod3

Compara estrategias de persistencia y motores de datos según consistencia, concurrencia, volumen, recuperación y operación.

mianhtsys/landingMod3

Design robust, scalable database schemas for SQL and NoSQL databases. Provides normalization guidelines, indexing strategies, migration patterns, constraint design, and performance optimization. Ensures data integrity, query performance, and maintainable data models.

mianhtsys/landingMod3

Consolida análisis especializados en una recomendación arquitectónica, matriz de decisión, ADR y plan de validación.

mianhtsys/landingMod3

Master error handling patterns across languages including exceptions, Result types, error propagation, and graceful degradation to build resilient applications. Use when implementing error handling, designing APIs, or improving application reliability.

mianhtsys/landingMod3

Build and maintain web frontends — component architecture, state management, API integration, responsive layout, client-side performance, and frontend testing patterns. Framework agnostic, focused on web frontend implementation. Do not use for backend service implementation, data engineering, or platform infrastructure work.

mianhtsys/landingMod3

Evalúa despliegue, contenedores, escalado, alta disponibilidad, observabilidad, backup, recuperación y CI/CD de forma proporcional.

mianhtsys/landingMod3

Execute read-only SQL queries against multiple Microsoft SQL Server databases. Use when: (1) querying MSSQL/SQL Server databases, (2) exploring database schemas/tables, (3) running SELECT queries for data analysis, (4) checking database contents. Supports multiple database connections with descriptions for intelligent auto-selection. Blocks all write operations (INSERT, UPDATE, DELETE, DROP, etc.) for safety.

mianhtsys/landingMod3

Generate and maintain OpenAPI 3.1 specifications from code, design-first specs, and validation patterns. Use when creating API documentation, generating SDKs, or ensuring API contract compliance.

mianhtsys/landingMod3

PHP 8.0+ development — XAMPP, RESTful APIs, PDO/MySQL/MariaDB, and authentication. Use when building PHP backends, creating API endpoints, configuring XAMPP, or integrating PHP with databases.

mianhtsys/landingMod3

Use this skill when designing or reviewing a PostgreSQL-specific schema. Covers best-practices, data types, indexing, constraints, performance patterns, and advanced features

mianhtsys/landingMod3

Analiza requisitos funcionales y no funcionales, reglas, restricciones, capacidad y ambigüedades antes de seleccionar tecnologías.

mianhtsys/landingMod3

Implement modern responsive layouts using container queries, fluid typography, CSS Grid, and mobile-first breakpoint strategies. Use when building adaptive interfaces, implementing fluid layouts, or creating component-level responsive behavior.

mianhtsys/landingMod3

>- Use when designing or implementing software securely: define security requirements, threat-model a feature, choose secure defaults, design authentication and authorization, handle untrusted data and secrets, evaluate dependencies, design multi-tenant trust boundaries, or review security-sensitive changes. Use for prevention during requirements, design, implementation, and review; not for post-build security assessments or scanning an existing codebase.

関連スキル